Microsoft continues to make progress on an issue that has become central to brands, publishers, and GEO teams: how to measure visibility in AI-generated responses.
With the evolution of Bing, Microsoft Copilot, and AI-enhanced search experiences, SEO is no longer limited to traditional rankings on search results pages. A brand can now be visible because it is mentioned, cited as a source, included in a generated response, or associated with a specific topic in a conversational experience.
That is precisely what Microsoft is aiming to make more measurable with the new features in the AI Performance report in Bing Webmaster Tools.
Four New Perspectives on Understanding AI Visibility
Microsoft is adding four new features to “Bing Webmaster Tools“: Intents, Topics, Citation Share, and Compare.
These new features do not replace traditional SEO metrics. They add a layerof analysis specific to visibility in AI-generated responses.
The goal is to shift to a more strategic analysis: Why was this content cited? In what context? On what topics? How much visibility did it receive? And how has it changed over time?
This is a significant development, because AI engines don’t just rely on isolated keywords. They interpret intentions, themes, sources, and credibility signals to generate a comprehensive response.
Intents: Understanding the Intent Behind Quotes
The first new feature, Intents, categorizes the queries or prompts that led to a quote into broad families of intent.
These intentions may be informational, commercial, navigational, local, search-oriented, learning-oriented, problem-solving, or content creation.
For GEO teams, the benefit is clear. It’s no longer enough to know that a page is cited; you have to understand the type of user journey in which it appears.
An e-commerce site, for example, might find that its content is primarily visible in AI experiences related to product comparison or purchase intent. An educational site, on the other hand, might see a strong presence in queries related to research or learning.
This intent-based analysis allows us to better tailor the structure, depth, format, and role of content within the user journey.
Topics: Moving from Keywords to Editorial Areas
The second feature, Topics, groups related queries into large thematic clusters.
This is a logical evolution. In an AI environment, systems do not rely solely on keyword matching. They organize information by concepts, themes, and semantic relationships.
For example, search terms such as “solar panels,” “solar energy efficiency,” or “residential solar installation” can be grouped under a single, broader theme related to solar energy.
For marketing teams, this approach is more in line with how a content strategy should be developed: by areas of expertise, customer pain points, and areas of interest, rather than by lists of isolated keywords.
This also sends a strong message to brands: subject-matter authority is becoming even more important. To be visible in AI-generated answers, it’s not enough to simply have a page optimized for a particular topic. You need to build a cohesive content ecosystem that demonstrates clear expertise in a given area.
Citation Share: Measuring Your Share of Visibility in AI Responses
The third new feature, Citation Share, is probably one of the most interesting.
This metric measures the proportion of citations attributed to a site for a given query, relative to the total number of citations displayed for that same query.
In other words, it’s not just a matter of whether your site appears. It’s about understanding where it stands within the AI-generated citation space.
This makes it possible to identify the topics on which a website has a strong presence, as well as those where its visibility is more fragmented or faces competition from many other sources.
Microsoft notes, however, that Citation Share should be understood as an observational metric. It is not a ranking score, a competitive ranking, or a traffic estimate. It does not reveal competing domains and should not be interpreted as a direct indicator of market share.
This is an important point: in AI-generated responses, a citation is not the same as a click. It measures a form of presence, credibility, or use as a source, but not necessarily a visit.
Compare: Track Changes Over Time
The fourth feature, Compare, allows you to compare the current period with a previous period.
Publishers can, for example, compare the last 30 days with the previous 30 days, or select custom time periods.
This feature is useful for tracking changes in AI visibility following a content update, a shift in demand, a seasonal effect, or a change in AI models.
It also helps us better understand that visibility in AI-generated responses is dynamic. Citations can vary depending on the recency of content, changes in models, new web signals, user behavior, or editorial competition.
What This Means for GEO
This announcement confirms a fundamental trend: GEO is entering a new phase of measurement.
For a long time, the key metrics were impressions, placements, clicks, and click-through rate. While these metrics remain important, they are no longer sufficient to understand a brand’s actual visibility in AI-driven environments.
Based on the responses generated, content can influence a decision without immediately generating a click. It can be cited as a source, contribute to a response, reinforce a brand’s credibility, or shape a user’s perception.
GEO reporting must therefore shift toward new metrics: visibility in AI-generated responses, presence by intent, topic authority, citation share, and trends over time.
Our Opinion
Bing Webmaster Tools has gained an interesting edge over Google here by offering metrics that are better suited to AI. Even though the data is still incomplete—particularly given the lack of click and click-through rate data—these new features already provide useful insights for guiding a GEO strategy. We’re now waiting to see how Google responds.
